DIMENSIONS OF NILPOTENT ALGEBRAS AMAJOR QUALIFYING PROJECT REPORT SUBMITTED TO THE FACULTY OF WORCESTER POLYTECHNIC INSTITUTE IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF BACHELOR OF SCIENCE WRITTEN BY KWABENA A. ADWETEWA-BADU APPROVED BY: MARCH 14, 2021 THIS REPORT REPRESENTS THE WORK OF WPI UNDERGRADUATE STUDENTS SUBMITTED TO THE FACULTY AS EVIDENCE OF COMPLETION OF A DEGREE REQUIREMENT. 1 Abstract In this thesis, we study algebras of nilpotent matrices. In the first chapter, we give complete proofs of some well-known results on the structure of a linear operator. In particular, we prove that every nilpotent linear transformation can be represented by a strictly upper triangular matrix. In Chapter 2, we generalise this result to algebras of nilpotent matrices. We prove the famous Lie-Kolchin theorem, which states that an algebra is nilpotent if and only if it is conjugate to an algebra of strictly upper triangular matrices. In Chapter 3 we introduce a family of algebras constructed from graphs and directed graphs. We characterise the directed graphs which give nilpotent algebras. To be precise, a digraph generates an algebra of nilpotency class d if and only if it is acyclic with no paths of length ≥ d. Finally, in Chapter 4, we give Jacobson’s proof of a Theorem of Schur on the maximal dimension of a subalgebra of Mn(k) of nilpotency class 2. We relate this result to problems in external graph theory, and give lower bounds on the dimension of subalgebras of nilpotency class d in Mn(k) for every pair of integers d and n. We conclude the thesis with some open problems. 2 Introduction A nilpotent matrix is a special case that, if multiplied by itself a certain number of times, equals the zero matrix. An algebra of nilpotent matrices is simply the operation of nilpotent, Addition, Multiplication, that allow for the result to still be nilpotent. In constructing algebras of nilpotent matrices, one might want to know of the limits to how small or how large the dimensions of such an algebra could be while still maintaining its properties. Using published resources, Mirzakhani and Erdös among others, as well as various literature in Linear Algebra and Graph Theory, this project will explore the upper and lower bounds on the dimensions of subspaces of nilpotent matrices in the matrix algebra Mn(k). For integers n,t we would like to compute the dimension of the largest subspace of Mn(k) which all matrices satisfy M t = 0. 3 Acknowledgements I would like to express my deep gratitude to Professor Ó Catháin for their guidance, encouragement, and infinite patience with this thesis. I would also like to thank Professor Weekes, and Guillermo Nunez Ponasso for their aid and resources, as well as the math department for their help in understanding the fundamentals that were the foundation of this thesis. Finally, I wish to thank my family for their support and backing throughout this endeavour. 4 CONTENTS 1. Linear Algebra and Jordan canonical form 5 1.1. Background 5 1.2. Diagonalisable matrices 8 1.3. Generalised Eigenvectors and generalised eigenspaces 10 1.4. Jordan Canonical Form 15 1.5. Nilpotent matrices 18 2. Algebras of nilpotent matrices 19 2.1. Definitions and elementary properties 19 2.2. The Lie-Kolchin theorem 21 3. Digraphs and associated matrices 23 3.1. Graphs, digraphs and their adjacency matrices 23 3.2. Matrix algebras associated with directed graphs 26 4. Theorems of Schur, Jacobson, Gallai-Hasse-Roy-Vitaver and Applications 28 4.1. The RREF basis of a nilpotent algebra 29 4.2. Algebras of nilpotency class 2 30 References 33 1. LINEAR ALGEBRA AND JORDAN CANONICAL FORM In this chapter, we review eigenvalues, eigenvectors and generalized eigenvectors of linear transformations. This material is essential for understanding the Jordan Canonical Form The- orem. Eventually, we will want to discuss the structure of a nilpotent linear operator. Then in Chapter 2, we will discuss algebras of nilpotent matrices. 1.1. Background. While we expect some familiarity with Linear Algebra, we will review all the notions required for the Jordan Canonical Form. While everything that we discuss holds over an arbitrary algebraically closed field, the reader may assume that all vector spaces are defined over C without issues. The material in this section is standard, and can be found in any Linear Algebra Textbook, for example Linear Algebra and its Application[3] and Linear Algebra, Fourth Edition[5] 1.1.1. Subspaces. Definition 1. ([5], 6-7) A vector space V over a field F consists of a set on which two operations, addition and scalar multiplication, are defined so that for each pair of elements x; y. in V there is a unique element x + y in V , and for each element a in F and each element x in V there is a unique element ax in V , such that the following conditions hold: (1) For all x; y in V , x + y = y + x (commutativity of addition). (2) For all x; y; z in V , (x + y) + z = x + (y + z) (associativity of addition). (3) There exists an element in V denoted by 0 such that x + 0 = x for each x in V . (4) For each element x in V there exists an element y in V such that x + y = 0. (5) For each element x in V , 1x = x. (6) For each pair of elements a, b in F and each element x in V, (ab)x — a(bx). (7) For each clement a in F and each pair of elements x; y in V , a(x + y) — ax + ay. (8) For each pair of elements a, b in F and each element x in V , (a + b) — ax + bx. Definition 2 ([5], page 16-17). A subset W of a vector space V over a field k is called a subspace of V if W is a vector space over k with the operations of addition and scalar multiplication defined on V . The zero subspace is also a subspace of V . Definition 3. ([3],196) Let v1; ··· ; vn be vectors in a vector space V . The Span of v1; ··· ; vn is n X Sp(v1; ··· ; vn) = αivijα 2 F i=1 Sp(v1; ··· ; vi) is always a subspace of V . If Sp(u1; ··· ; un) = Sp(v1; ··· ; vn), then u1; ··· ; un is a generating sset for Sp(v1; ··· ; vn) Definition 4. ([3], 211) Let H be a subspace of a vector space V . A set of vectors B = b1 ··· bn in V is a basis for H if: (1) B is a linearly independent set, and (2) the subspace spanned by B coincides with H ; that is, H=Span ( b1 ··· bn ) 6 Theorem 5 ([5], page 16-17). Let V be a vector space and W a subset of V. Then W is a subspace of V if and only if the following three conditions hold with the operations defined in V. (1) 0 2 W (2) x+y 2 W, whenever x 2 W and y 2 W (3) αx 2 W, whenever α 2 k and x 2 W 1.1.2. Linear transformations and matrices. Definition 6. Let V and W be Vector spaces. A function T : V −! W is linear if it holds to the following: (1): T (v1 +v2 ) = T (v1) + T (v2), for all v1; v2 2 V (2): T (αv)= αT (v), for any α 2 R, for all v1; v2 2 V Definition 7. If T be a linear transformation and B is a basis for V , the matrix M with respect to B is as shown below: B [M]B = [(Mb1)Bj(Mb2)Bj · · · j(Mbn)B] B with Mbi being interpreted as the image of bi under M .with that knowledge, [M]B can be written as shown below, 2 3 a1 6a27 a b + a b ··· a b = 6 7 1 1 2 2 n n 6 . 7 4 . 5 a n B this way that the matrices can be described in is in terms of change of basis, this way it could give future proving of proofs a more structured foundation onto which they could be attained. 1.1.3. Eigenvalues and eigenvectors. Definition 8. Let T be a linear operator on a vector space V . An eigenvector of T is vector v such that T v = λv for some scalar λ. We refer to λ as the eigenvalue of T associated with v. Suppose that we want to find the eigenvectors of a matrix M . To do this, We will be some initial definitions. a b Definition 9. ([3], 105) Let A= : The determinant of A = ad - bc c d Definition 10. ([5],209; [10], 396) Let M be a n × n matrix as shown as below: 2 3 m11 ··· m1n 6 . .. 7 M = 4 . 5 mn1 ··· mnn 7 The (n − 1) × (n − 1) matrix M i;j can be obtained by deleting the ith row and j th column as follows: 2 3 m1;1 ··· m1j−1 ··· m1;j+1 ··· m1n 6 . .. .. .. 7 6 . 7 6 7 6mi−1;1 ··· mi−1;j−1 ··· mi−1;j+1 ··· mi−1;n7 6 . 7 M i;j = 6 . .. .. .. 7 6 7 6m ··· m ··· m ··· m 7 6 i+1;1 i+1;j−1 i+1;j+1 i+1;n7 6 . 7 4 . .. .. .. 5 mn;1 ··· mn;j−1 ··· mn;j+1 ··· mn;n Definition 11. ([5], 209-210) Let M be a n × n matrix, with m ≥ 2. the determinant of M is as follows: n X 1+j det(M) = (−1) × M1j × det(M i;j) : j=1 Theorem 12.